fix: unavailable model on startup

app/helpers/_modelclients.py

@@ -9,7 +9,7 @@
 from app.schemas.embeddings import Embeddings
 from app.schemas.models import Model, Models
 from app.utils.config import LOGGER
-from app.utils.exceptions import ContextLengthExceededException, ModelNotFoundException, ModelNotAvailableException
+from app.utils.exceptions import ContextLengthExceededException, ModelNotAvailableException, ModelNotFoundException
 from app.utils.variables import EMBEDDINGS_MODEL_TYPE, LANGUAGE_MODEL_TYPE
 
 
@@ -66,12 +66,20 @@ def get_models_list(self, *args, **kwargs):
 def check_context_length(self, messages: List[Dict[str, str]], add_special_tokens: bool = True):
     headers = {"Authorization": f"Bearer {self.api_key}"}
     prompt = "\n".join([message["role"] + ": " + message["content"] for message in messages])
-    data = {"model": self.id, "prompt": prompt, "add_special_tokens": add_special_tokens}
+
+    if self.type == LANGUAGE_MODEL_TYPE:
+        data = {"model": self.id, "prompt": prompt, "add_special_tokens": add_special_tokens}
+    elif self.type == EMBEDDINGS_MODEL_TYPE:
+        data = {"inputs": prompt, "add_special_tokens": add_special_tokens}
 
     response = requests.post(str(self.base_url).replace("/v1/", "/tokenize"), json=data, headers=headers)
     response.raise_for_status()
+    response = response.json()
 
-    return response.json()["count"] <= response.json()["max_model_len"]
+    if self.type == LANGUAGE_MODEL_TYPE:
+        return response["count"] <= response["max_model_len"]
+    elif self.type == EMBEDDINGS_MODEL_TYPE:
+        return len(response[0]) <= self.max_model_len
 
 
 def create_embeddings(self, *args, **kwargs):
@@ -91,17 +99,20 @@ def create_embeddings(self, *args, **kwargs):
 class ModelClient(OpenAI):
     DEFAULT_TIMEOUT = 10
 
-    def __init__(self, type=Literal[EMBEDDINGS_MODEL_TYPE, LANGUAGE_MODEL_TYPE], search_internet: bool = False, *args, **kwargs):
+    def __init__(self, type=Literal[EMBEDDINGS_MODEL_TYPE, LANGUAGE_MODEL_TYPE], *args, **kwargs):
         """
         ModelClient class extends OpenAI class to support custom methods.
         """
         super().__init__(timeout=self.DEFAULT_TIMEOUT, *args, **kwargs)
         self.type = type
-        self.id = None
-        self.owned_by = None
-        self.created = None
+
+        # set attributes for unavailable models
+        self.id = ""
+        self.owned_by = ""
+        self.created = round(time.time())
         self.max_model_len = None
-        self.search_internet = search_internet
+
+        # set real attributes if model is available
         self.models.list = partial(get_models_list, self)
         response = self.models.list()
 
@@ -110,9 +121,7 @@ def __init__(self, type=Literal[EMBEDDINGS_MODEL_TYPE, LANGUAGE_MODEL_TYPE], sea
             self.vector_size = len(response.data[0].embedding)
             self.embeddings.create = partial(create_embeddings, self)
 
-        # @ TODO : extends to embeddings models
-        if self.type == LANGUAGE_MODEL_TYPE:
-            self.check_context_length = partial(check_context_length, self)
+        self.check_context_length = partial(check_context_length, self)
 
 
 class ModelClients(dict):
@@ -121,16 +130,16 @@ class ModelClients(dict):
     """
 
     def __init__(self, config: Config):
-        for model in config.models:
-            model = ModelClient(base_url=model.url, api_key=model.key, type=model.type, search_internet=model.search_internet)
+        for model_config in config.models:
+            model = ModelClient(base_url=model_config.url, api_key=model_config.key, type=model_config.type)
             if model.status == "unavailable":
-                LOGGER.info(f"error to request the model API on {model.url}, skipping.")
+                LOGGER.error(f"unavailable model API on {model_config.url}, skipping.")
                 continue
             self.__setitem__(model.id, model)
 
-            if model.search_internet and model.type == EMBEDDINGS_MODEL_TYPE:
+            if model_config.search_internet and model_config.type == EMBEDDINGS_MODEL_TYPE:
                 self.SEARCH_INTERNET_EMBEDDINGS_MODEL_ID = model.id
-            if model.search_internet and model.type == LANGUAGE_MODEL_TYPE:
+            if model_config.search_internet and model_config.type == LANGUAGE_MODEL_TYPE:
                 self.SEARCH_INTERNET_LANGUAGE_MODEL_ID = model.id
 
         if "SEARCH_INTERNET_EMBEDDINGS_MODEL_ID" not in self.__dict__: